Slashdot Mirror
Nvidia Uses AI To Render Virtual Worlds In Real Time (tomshardware.com)
Nvidia is using artificial intelligence to draw new worlds without using traditional modeling techniques or graphics rendering engines. "This new technology uses an AI deep neural network to analyze existing videos and then apply the visual elements to new 3D environments," reports Tom's Hardware. From the report: Nvidia claims this new technology could provide a revolutionary step forward in creating 3D worlds because the AI models are trained from video to automatically render buildings, trees, vehicles, and objects into new 3D worlds, instead of requiring the normal painstaking process of modeling the scene elements. But the project is still a work in progress. As we can see from [this image], which was generated in real time on a Nvidia Titan V graphics card using its Tensor cores, the rendered scene isn't as crisp as we would expect in real life, and it isn't as clear as we would expect with a normal modeled scene in a 3D environment. However, the result is much more impressive when we see the real-time output in [this YouTube video]. The key here is speed: The AI generates these scenes in real time.
Whenever I see someone creating things like this it makes me realise how dumb I am in comparison! :-)
You folks dont exist until I go on slashdot... whos cat was that :)
[($)]
Its like the deepfakes. Make a movie of anyone doing anything.
Real time.
For a 100 year old oak, it only needs a hundred years, real time.
They use AI *and* Unreal Engine 4 to render virtual worlds.
Thirty four characters live here.
It seems more like the AI is procedurally generating the world, while the actual rendering is still using the Unreal 4 engine pipeline. It is impressive, but the title /is/ misleading.
Right now they apparently feed the AI video and it creates models and animations based off that video, which it then recombines into new worlds. Imagine this tech combined with GoogleEarth's library of images: it could create a fully-3D view of every street. It will be sort of like jumping from the old step-based Bards Tale games to Skyrim.
...was pretty damned impressive. Training sims will be a lot more effective anyway.
Social Media Handywoman at Texas Boys Balloo
More specifically, they use AI for filling polygons rendered by UE4.
What the fuck? Not even close.
If successful it would do little more rob people of a creative outlet and a living, for a lot of that stuff you saw in computer games was art, as well as bread-and-butter. What problem does it solve?
https://www.youtube.com/c/BrendaEM
Waiting for someone to comment "that's not AI", as always happens with any story about AI.